Video problems / Questions

[sonicmook56]

Here at the studio, I found it annoying to switch inputs to a video monitor and external console inputs, between a DVD player and a satellite recever.

First you had to patch the video, then patch the audio, then hit an external select switch on the console.

I came up with this little box that takes the unalanced audio output and ballances it, as well as take the audio and video and switches it via toggle switch.

Before I explain my video problem, I just want to say that I have never worked, or even thought about a video circuit. The circuit in question is straight from the Mfg. demo board layout.

http://datasheets.maxim-ic.com/en/ds/MAX4313EVKIT.pdf

So I decided to go with a MAX4313 to switch the video. (still not sure why this IC is better than an simple switch) Testing difftent sources with this gives me diffrent results on the output... I'm sure I'm missing somthing fundamental

Satallite recever: Looks good on the monitor
DVD: hardly anything, black and white all distorted. If I remove one of the 49.9ohm resistors, I get a picture, but it's very washed out "white"

I tested it on another DVD player and had the same results.

I looked at the NTSC on my scope and noticed a diffrent offest on the good signal, compared to the bad signal..

I checked the DC offset of the satallite recever, and it was around .6Vdc
The DVD player in question was at 0v

Could this be an input bias problem, input capactance?

Again, I'm sure I'm missing somthing fundamental here and I have no idea what I'm doing video wise.

The audio part of the circut measures .00063 thd. Nice and clean.




:sam: :sam: :sam: :sam:

 

[Walrus]

Isn't composite video 75 Ohm? And that circuit looks like it's designed for 50 Ohm. But I don't think that's what is affecting the DVD player, it sound more like the DVD is outputting S-Video and you are just seeing the brightness signal. Try checking the DVD menu settings.

 

[rodabod]

Yes, first things first - are you actually getting a composite signal from the DVD player?

 

[NewYorkDave]

Yes, R1, R2 and R3 should be changed to 75 ohms.

What happens if you swap the inputs? Does the distortion follow the signal or the switcher input? If you connect the DVD directly to the monitor, the display is good?

 

[burdij]

You need a sync tip clamp. Video circuits are very sensitive to the DC reference of the video signal which occurs in the area between the rising edge of the horizontal sync pulse and the start of the color burst. This point is usually clamped to ground during every horizontal line. An almost as effective method is the sync tip clamp. This clamps the most negative portion of the horizontal sync pulse to a reference. There aren't any really simple and effective sync tip clamps but MAXIM makes an output buffer with an input sync tip clamp that you could place after your switch. The part number is MAX4090. It also provides 6db of gain to overcome the losses that occur due to cable terminations. Caps in video circuits contribute to the ground reference problem and are not your friends. BTW - video cables are terminated in 75 Ohms so you will need to change those 49 Ohm resistors.

Here is a link to the MAXIM datasheet:

http://datasheets.maxim-ic.com/en/ds/MAX4090-MAX4090A.pdf

Here is a link to a diagram showing the NTSC signal structure:

http://www.sxlist.com/techref/io/video/ntsc.htm

 

[rodabod]

I guessed the video monitor would have been capable of clamping the signal. What we be the cause of the problem then - I thought the ground reference would still be maintained in this switch.

 

[burdij]

Yes it would if the capacitor on the output of the DVD which is causing the problem was directly connected to the monitor. I did notice that the MAXIM switch specifies a minimum voltage input of 0V. During the vertical retrace interval of the video signal, the capacitor in the DVD will probably cause the signal to drift more negative than 0V. This may be causing the overall problem. The switch is not passing the lower parts of the video signal from the DVD, the part containing all the sync information and the color burst. It might be possible to trick the switch by biasing the DVD output with a voltage and then passively clamping the sync with a diode to raise the overall waveform above 0V.

 

[rodabod]

Ah, I see. Thanks.

[quote author="burdij"] I did notice that the MAXIM switch specifies a minimum voltage input of 0V.[/quote]

That's a bit crap! Your idea of biasing a little might be a good idea to solve that problem though.

 

[moamps]

[quote author="sonicmook56"]
Could this be an input bias problem, input capactance?
[/quote]

Yes,
just pull Vee pin of the MAX to -5V.
Regards,
Milan

 

[NewYorkDave]

Yeah, that is pretty bad for a "video switching" IC. A video signal with proper DC level, by its very nature, goes negative as well as positive.

It's too late for the original poster, but I have implemented simple two-way video switches successfully with Omron 900MHz relays. It's important to pay attention to layout and also to ensure that the "unselected" input is terminated to reduce the likelihood of crosstalk.

Of course, you can find multi-input audio-video selectors for cheap at any Best Buy, Circuit City or Rat Shack and these are generally adequate for non-critical use.

 

[sonicmook56]

Yes, the signal is definatly NTSC, composite. If I connect the cable directly to the monitor it works fine.

I changed all the 49.9 resistors to 75 ohms and added an extra 75 ohm on the output to ground.

This seems to have helped as the DVD now works (1/2 the time) but looks like it might spiral out of control at any moment. Satallite input is just fine. Problem follows the output of the DVD, not the input to the switch.


Here are some shots from my scope, The first one being blackburst.

This one is the output of the satallite recever (speed channel)

and the splash screen of the DVD player.

As NYD said, the video goes negative, and I'm running the MAX4313 on a single ended supply (5v) I bet if I was using a bi-polar supply like moamps said, this would work ok as the signal could swing below gnd (witch it's obvousley not doing)

It's too late for me to re-design the board, so I might just super glue little relays in place of the MAX ic and be done with it.

Are these the correct ones to use for this?

http://oeiwcsnts1.omron.com/pdfcatal.nsf/PDFLookupByUniqueID/4D29F4FD245FAB4E86256D350052C09A/$File/D20G6Y0503.pdf?OpenElement

 

[sonicmook56]

Ok, so I just super glued a TO-92 7905, a couple of caps and pulled pin 6 , Vee to -5V and it works perfectly now.

You guys are the best.

~B

 

[NewYorkDave]

Why the "extra" 75 ohm to ground on the output? When you plug it into your monitor, the output will be double-terminated and you'll lose half your signal amplitude.

Even if the monitor has a bridging "loop-through" input, it's better to terminate the line at the destination.

 

[sonicmook56]

oh, I took that out after adding the negitive vReg. I don't rember why I thought that should be in there.

This box works awsome. I'm very happy with the way it turned out. Just need to get my CNC working so I can have a nice front panel. :sam: